I'm on my hols, and this holiday is dog training time - although I do still seem to have a ridiculous amount of random 'stuff' to do . I've been super short of time for a good 2 months now, and apart from trying to move my 'proper' heel on a bit during walks (as opposed to wander and wobble round me like an over loaded ant carrying a big leaf but just don't pull ) I haven't done much. So out on the extremely muddy field today, to take stock of where we are - not tooooo bad, on the whole. But because I was running through quite a few things in short order as a bit of a stock take, Charlie's addiction to patterns was SO obvious. I work hard on putting things on verbal cue, and I really know how, but patterns can really overwhelm my cues and I'm just wondering whether, or how, I can use this to my advantage. Charlie generalises extremely poorly, which is why he is such hard work at times, but he learns a pattern in a flash - he gets an idea in his head "I know! It's X next, because that's what happened last time!" and he's off...even though it was Y next...... How can I use this to my advantage, I wonder? I only have to repeat a sequence a couple of times and then Charlie is 'in that groove', goes deaf to his cues, and I struggle to get him out of it. This is one of the main reasons I find him a difficult dog (I do have to hold my hand up and admit he might be a total doodle of a dog, and I', hopeless, but nevertheless I find him difficult and I do have a fair number of comparators).
So just a couple from this afternoon.... I did a couple of retrieves, and then did a sit and wait while I walked away. I didn't put anything down, it was just my normal - done it a thousand times - "wait" and I walk away and walk back. I tell Charlie 'close' (heel) and he dashes out as though I've said 'fetch'. Because the previous two times I'd said fetch. Both fetch and close have had a lot of work on them in terms of cue discrimination. I walk along the field polishing up my stop whistle on a walking - walk at heel, blow stop, keeping walking, say close to return the dog to heel. Did it twice. Then for the rest of the circuit of the field, Charlie keeps stopping, randomly, while at heel even when I'm not blowing my stop.
Hmmm, I can't think of how to use it to your advantage, sadly; hopefully someone more experienced will help you out with that. I can only offer you SWMBO's phrase of "anticipate anticipation'. Obviously, Charlie is a super smart cookie and works patterns out really quickly, which, I suppose, means you have to either make things different each and every time, or teach him not to anticipate by withholding the rewards, which I know is something he has a strong reaction to, so needs approaching with care so he doesn't become frustrated. Maybe, where you have done X followed by Y, you then just ask for X and treat heavily, at first potentially bribing him to stay in place, gradually extending the time between the behaviour and the treat, so he learns to stick with you until you cue Y, Z or whatever. This would be similar to when learning stop at a distance, where you have to ensure that you ask for and reward a straight run to the target far more frequently than you do the stop, so the dog doesn't lose its zip, anticipating the stop. But you know all this anyway!
It should be really powerful, I think. I just need to work out where his inclination to repeat something will overcome a distraction....without not being able to get him to change later. I think if I did the same thing in exactly the same way, more times than I normally would (because then he'd get stuck) I can introduce distractions faster....
could you use it to set up a "hunt - recall/retrieve" pattern ie every hunt is followed by a recall or retrieve or would Charlie need it to be as precise as a 10 sec hunt/retrieve. The first could be useful - the 2nd might not be - sorry just throwing ideas in
Ideas very welcome - I often think with Charlie I have to use what he is inclined to do. Hunting is a strange one with Charlie. I have great difficulty stopping him searching once he starts. Yesterday when he ran out as though I'd sent him on a blind (when I'd said 'close') he was a bit hesitant (as he bloomin' well should have been since he had no reason whatsoever to think there was a dummy out) and I did manage to stop him. As soon as I'd stopped him, I threw a ball. So even though it was all a complete accident, it did turn out to be useful as I got a stop on a search, which I've struggled with (as soon as I get the behaviour I can reward with a ball and then I'll get it again but I was struggling to get it at all).
Wow interesting? He either has not learnt the meaning of the two cues, or you are not giving them as a definate cue. Basically he should not go for the retrieve until you send him? How are you sending him for a retrieve? On a verbal cue (fetch) or are you getting down and pointing in the direction of the dummy and the verbal (fetch) We use 'Go back' (but that does not matter). If a dog is extremely keen on dummies, which I sense Charlie is, Any verbal word maybe set him off. ( a cough may send some dogs ). The way we train is to go down and point to the dummy and give the verbal command (cue), on the odd occassion go down and point and not give the cue and stand up again, so the dog does not learn that going down and pointing is the cue to go. Close or Heel with a tap on the thigh should be enough for the dog stay by your side. If you have not done it for a couple of months, he may need a refresher.
Charlie is very intelligent and very keen to follow his cues and be rewarded, unfortunately he also anticipates what he thinks you are going to want next , one of the down sides of an intelligent dog. I know Juno is no where near as well trained as Charlie and is probably a bit more laid back but with her if she anticipates I tend to ignore what she is doing and just move on and ask her for something else instead. I tend not to do repetitions of things in a row and mix them up so she has to focus on me. Not sure if that approach would work with Charlie though.
What Charlie is, is a mad "doer" of a dog, and a bit bonkers with it - although you are all very kind to say he's intelligent (I'm sure he is, of course).
Very intelligent and ever hopeful! He sounds like he'd be brilliant at agility! Maybe teach patterns, with a clear cue as to which pattern it will be, followed by the treat at the end? My training is very, very simple in comparison but I'm very random as to when they get a treat for an action and when they don't.. Sometimes it's a huge jackpot for a simple command, other times it's nothing, other times it's a bit of kibble. Until they know a command they get a treat every time, of course. Then I have a complicated method of fading treats (which I write down!) then I do it randomly. I always have my hand behind my back for the signal so they don't know if a treat is there or not. At home I put treats ready in different places so I don't always get it from the smae place. So they always hope, of course, but they don't anticipate.
I should think Charlie would be lousy at agility. He'd be a whining, over excited, horror, unable to think straight let alone remember if he was supposed to go under or over something. Yes, I've started clicking for not doing an action after I saw that at clicker expo - it does work, not fantastically well, but it definitely helps. I've never had to do less than two of something in order to not have Charlie anticipate a repetition before, doing just one of anything will be somewhat limiting. But anyway, avoiding anticipation is standard (if somewhat painful with Charlie) - it's interesting to think about how to use this though, I mean lots of clicker based repetitive exercises harness the dog's expectation to repeat the last thing that got him a treat, so just thinking more broadly....
An interesting thought that Juno sometimes gets a treat, sometimes she doesn't and the value of the treat is random as well, rather like Mags but she hasn't been clicker trained so is her expectation/anticipation different? She gets a "good girl" and a treat when we're doing something new and I may then repeat a couple of times and treat for each and repeat during the day at odd times. Once we have a consistent approach I reduce the treats but we always have a "good girl"
Well, I don't know what Juno's anticipation is like...although I wouldn't have thought a reward schedule impacts on anticipation, it's more about the predictability of the trainer, I think. And the personality of the dog/willingness to work is a big factor I would have thought.
It doesn't have to be just one of something, just one of something in a particular order. So, for example, in your first example, sit and walk away, ask for a fetch. Next time, sit and walk away, ask for a down then a fetch. Next time, sit and ask for a hand target, then sit and walk away, walk back, ask for a close, walk away, ask for a fetch.... you're still getting your fetch in, assuming this is the thing you're working on, but the chain is different each time.
About the only time Juno anticipates is when I've put her in a sit and walked away from her when we're out on a walk. It has been known that I turn around and she's about 3 feet behind me with her tail and body wagging . I just walk on and try again further down the path and then she does sit and wait for her cue. She is a very calm, quiet girl (I sometimes wonder if she really is a Lab at all ) and is happy to work when asked although being honest she has a lazy trainer
Back to the original question, which was not avoiding anticipation, but whether it could be used in some way? But thanks for all the tips all the same....
Well, I guess it's useful inasmuch as he's quickly learning chains of behaviours. So, if there are chains you want to build, you're going to have an easier time of it than people whose dogs aren't so quick to work out patterns. Maybe for some Premack, too, as long as you can interrupt the pattern, but I have a feeling that that's probably easier said than done with Charlie!
I was wondering about the pros (and cons) of drills. I mean, deliberately repeating something over and over....Pippa describes its purpose as 'fixing a concept' in the dog's mind.
